Re: Unable to disable IDE DMA on boot

From: Agustin Martin (agmartin_at_debian.org)
Date: 03/31/04

  • Next message: Torbjörn Schultz: "X -window how can I start it?"
    Date: Wed, 31 Mar 2004 16:51:18 +0200
    To: debian-user@lists.debian.org
    
    

    Hi, Johannes,

    I had to deal with this problem too and got some experience on it

    > Then I tried to create a custom initrd image using "mkinitrd" from
    > "initrd-tools" (0.1.56). I noticed that the default "/sbin/init" script
    > will try to pass the IDE options specified on the kernel command line to
    > "ide-mod" module. However, this module has been renamed to "ide-core" so

    That will be fixed in one of initrd-tools next uploads (see #240734). In
    the meantime you can use

    alias ide-mod ide-core

    but I suggest you to do the thing below. By the way, even with a fixed
    initrd-tools there is a problem about how 2.6 kernel passes options to
    the modules (see #240886 for details) so there will be problems with 2.6.

    > Then I added into "/etc/mkinitrd/modules" the line (also tried other
    > variations above)
    >
    > ide-core ide=nodma

    ide-core options="ide=nodma"

    This is the expected format (see ide.txt) and this is the way I suggest
    you (ide kernel command line boot options are grouped together by
    initrd-tools and prefixed by options). This will not work in 2.6 if more
    than one ide option is there, because of the above mentioned bug, but
    will casually work if is just as above.

    > I then created the new custom initrd image using just "mkinitrd -o
    > /boot/myinitrd.img", updated the "/initrd.img" link and ran lilo.
    > However, this did not help at all. The DMA was still enabled.

    I did my tests keeping a safe boot image with DMA disabled from which I
    can always boot from, and doing my tests in an initrd image having DMA
    enabled by default. To make sure I am doing exactly the same that is
    done at package installation I reconfigured the kernel-image package, e.g.

    # dpkg-reconfigure kernel-image-2.4.25-1-586tsc

    although if you are building a initrd for your running kernel your call
    should work.

    Hope this helps

    -- 
    Agustin
    -- 
    To UNSUBSCRIBE, email to debian-user-request@lists.debian.org 
    with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
    

  • Next message: Torbjörn Schultz: "X -window how can I start it?"

    Relevant Pages

    • Re: [PATCH][TAKE 3] THE LINUX/I386 BOOT PROTOCOL - Breaking the 256 limit
      ... A string that is too long ... -will be automatically truncated by the kernel, a boot loader may allow ... +The kernel command line is a null-terminated string. ...
      (Linux-Kernel)
    • [Bug 1430] New: SysFS oops when rmmoding uhci-hcd after resuming from suspend
      ... "init=/bin/bash" kernel command line parameter. ... Hardware Environment: ... Unable to handle kernel NULL pointer dereference at virtual address ... Steps to reproduce: ...
      (Linux-Kernel)
    • Re: root=/dev/sda1 fails but root=0x0801 works...
      ... if I have my kernel command line with ... SCSI device sda: 2001888 512-byte hdwr sectors ... Typically lilo will translate /dev/sda1 to 0x0801 automatically for you and pass that to the kernel for this reason. ... These days most distributions are using an initrd which the kernel mounts as the root fs, and that sets up a minimal /dev and does some hardware detection. ...
      (Linux-Kernel)
    • Re: [BUG] local_softirq_pending storm
      ... If I disable NOHZ by kernel command line nohz=off this problem ... This system is P4/2.40GHz/HT with SMP/SMT on in kernel config. ... network otherwise plain command line with no network access don't trigger ... The patch below ratelimits the printk output, ...
      (Linux-Kernel)
    • [PATCH] THE LINUX/I386 BOOT PROTOCOL - Breaking the 256 limit - Resubmit
      ... Extending the kernel parameters to a size of 1024 on boot ... This patch was submitted a long ago, but did not find its way into the kernel, but was not rejected as well. ... architectures, please reply. ... +The kernel command line is a null-terminated string. ...
      (Linux-Kernel)